JPH0744638A - Table data retrieving device - Google Patents

Abstract

PURPOSE:To cope with a problem that the burden for a user is heavy when data recorded on spread sheet are retrieved from a data base by realizing a table retrieving device by which the data base retrieval from spread sheet is easy. CONSTITUTION:This device is composed of a table means 001, a natural language interface means 200, a data base 300, an explanatory sentence input means 101 and a retrieval result extraction means 191. The explanatory sentence means 101 receives the explanatory sentence 901 of a natural language on the data recorded on the table means 001 from a user and inputs the sentence in the natural language interface means 200. The explanatory sentence 901 is allowable even if it is a retrieval instruction sentence for retrieving the data recorded on the table means 001 from the data base 300. The natural language interface means 200 inputs the explanatory sentence 901 by the natural language and outputs a retrieval expression 902 for retrieving the data base 300.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of inputting data used in a spreadsheet for tabulating data, and particularly for obtaining data to be input to the spreadsheet by searching a database existing independently of the spreadsheet. Table data retrieval method.

[0002]

2. Description of the Related Art Conventionally, there has been a method of inputting data from a database to a spreadsheet by directly reading a file storing the data of the database and inputting it into a table. In order to directly read the file that stores the data, it is necessary to clarify the structure of the database and the format of the file that stores the beta.

Further, in order to search from a database input in a table, a search formula for the database is input and a database search is executed. With this method as well, it is necessary to describe the search formula for database search after the structure of the database has been clarified.

[0004]

With the conventional method of directly reading a database file, it is difficult to selectively read data because the table format at the time of reading is restricted by the database storage format. Later, the user had to format the table into the desired format.

Further, in the method of inputting a search formula, it is necessary for the user to be able to describe the search formula, and it is necessary to acquire the method of describing the database search formula in addition to the method of using the spreadsheet.

Further, in any of the methods, there is a problem that the data cannot be used on the spreadsheet unless the details of the database are well known. Therefore, in order to use the data of the database on the tabulation spreadsheet. Has a problem that the user's burden is heavy.

An object of the present invention is to accept an input of a search sentence in a natural language from a spreadsheet, to execute a database search via a natural language interface for a database, in order to facilitate a database search from a spreadsheet. It is to provide a tabular data search method that enables input of a search result into a spreadsheet as tabular data.

[0008]

The gist of the means for solving the above problems is in the following three items.

[1] In a tabular data retrieval method for a spreadsheet having table means for recording data in a tabular form, an explanatory text input means for inputting an explanatory text relating to the data to be recorded in the table means in a natural language, Natural language interface means for converting the inputted explanation sentence into a command understandable by the database and generating a database search expression, and performing a database search by the database search expression, and transferring the result of the database search to the table means. A table data search method comprising a search result extracting means.

[2] It is assumed that the table means has a title column for recording the item name corresponding to the data, the item name recorded in the title column is input, and when there are a plurality of input item names, these items are input. The table according to claim 1, further comprising: an explanatory sentence generating unit that generates an explanatory sentence regarding the data by combining the names, and the explanatory sentence generated by the explanatory sentence generating unit is input to the natural language interface unit. Data search method.

[3] The explanation sentence recording means for recording the explanation sentence corresponding to the data recorded in the table means, and the explanation sentence recorded in the explanation sentence recording means are input, and the input explanation sentence is plural. And a description sentence synthesizing means for generating a synthesis explanation sentence in which these explanation sentences are combined, and the synthesis explanation sentence generated by the explanation sentence synthesizing means is input to the natural language interface means. The table data search method according to item 1.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the table data search method of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a table data search method in an embodiment of the invention of claim 1.

The table data retrieval method in this embodiment is as follows:
Display means 001 and natural language interface means 200
A database 300, an explanation input means 101,
The search result extracting means 191 is included.

The table data retrieval method of claim 1 will be described with reference to FIG. The explanatory text input means 101 is a natural language explanatory text 90 regarding the data to be recorded in the display means 001.
1 is received from the user and input to the natural language interface means 200. The explanatory note 901 may be a search command statement for searching the database 300 for data to be recorded in the table means 001.

The natural language interface means 200 can use, for example, the natural language interface method described in Japanese Patent Application No. 4-256660 filed by the applicant of the present invention, and inputs a description 901 in a natural language such as Japanese, A search formula for searching the database 300 is generated.

For example, if the database 300 is a relational database and accepts SQL that is widely used as a search language for the relational database, the natural language interface means 200 is, for example, the IPSJ 43rd National Convention Lecture Collection 3- SQ configured to include the search expression generation method reported on pages 175 to 3-176.
The search expression 902 described in L is output.

The search result extracting means 191 is the database 3
If the search result 911 output from 00 includes information other than the data such as the number of searched data other than the one recorded in the table means 001, the search result 911 may be deleted or the search result 911 may be deleted. When the delimiter of the data 911 is different from the one recognized by the table means 001, the delimiter is converted and the search result 9 is obtained.
The table data 912 obtained by shaping 11 is recorded in the table means 001.

In FIG. 5, assuming that the database 300 contains information about company names and addresses,
An example in which a sentence “name and address of a company in M city” is assumed to be input as the explanatory sentence 901 relating to the table is shown. From the search result 911, the table data 912 obtained by converting a semicolon into a colon is input to the table means 001 except for the top line "There are 3 cases." When the table means 001 recognizes a colon as a delimiter, the table 001 is obtained.

FIG. 2 is a block diagram showing the structure of a table data searching method in the embodiment of the invention as claimed in claim 2. The table data search method of claim 2 will be described with reference to FIG.

In this embodiment, the tabular data search method is a tabular means 001 having a title column 003, an explanation sentence generating means 102, and a natural language interface means 200.
And a database 300 and a search result extracting means 191.

A title column 003 is set in the display means 001. The title column 003 is set to only columns as in the embodiment of FIG. 6, only rows as in the example shown in FIG. 7, or both rows and columns as in the example shown in FIG. It may be set. Further, the title columns of rows and columns may include item names 903 hierarchically.

FIG. 9 shows an embodiment in which the column item names 903 form two layers.

The explanation sentence generation means 102 uses the title column 003.
The item information 905 recorded in is input, and a description 906 for searching the database 300 for data corresponding to the item information 905 is output. The item information 905 includes an item name 903 and position information 900.

From the example of the title column shown in FIG.
For example, the sentence shown in FIG. 11 is generated. In the embodiment of FIG. 8, one of the item names 903 in the row title column and one of the item names 903 in the column title column is used as the search target, and the other is used as the search condition to generate the sentence shown in FIG. 12, for example. Also,
For example, when it is desired to perform a database search only on a specific area 801 in the table, for example, a sentence shown in FIG. 13 is generated from the item names 903 “profit” and “second half” related to the area 801. When the item names 903 are hierarchical as in the example shown in FIG. 9, the sentence shown in FIG. 14, for example, is generated by the conjunction of the hierarchical item names 903. The sentence generation from the item name 903 can be generated by a generation rule as shown in FIG. 10, for example.

Description 906 generated from item name 903
Is a search expression 9 by the natural language interface means 200.
It is converted into 07 and input to the database 300. The search result 913 based on the search expression 907 is the search result extraction means 19
The data is shaped into the table data 914 by 1 and is recorded in the position corresponding to the item information 905 in the table means 001.

FIG. 3 is a block diagram showing the structure of a table data searching method according to the third embodiment of the invention. The table data search method of claim 3 will be described with reference to FIG.

The table data search method in this embodiment is configured by adding a description sentence recording means 103 and a description sentence synthesis means 104 to the table data search method which is an embodiment of the invention of claim 1.

The description storage means 103 has a display means 001.
Record one or more explanations corresponding to the data recorded in. The explanation sentence synthesizing unit 104 receives one or more explanation sentences 908 output from the explanation sentence recording unit 103, and outputs the explanation sentence 909 obtained by combining the explanation sentences 908. The explanatory sentence 909 which is the output of the explanatory sentence synthesizing unit 104 is converted into a search expression 910 by the natural language interface unit 200 and input to the database 300. The search result 915 according to the search formula 910 is shaped into the table data 916 by the search result extracting means 191 and recorded in the table means 001.

Further, as shown in FIG. 4, the display means 001 may include a title column, and the explanatory note may be recorded in the explanatory note recording means 103 in association with the item name 903.

[0030]

According to the tabular data retrieval method of the present invention, an input of a retrieval sentence in a natural language is accepted from a spreadsheet, a database retrieval is executed via a natural language interface for a database, and the retrieval result is used as tabular data. It is possible to input data into a spreadsheet and search the database from the spreadsheet.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a table data search method according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a table data search method according to an embodiment of the invention of claim 2;

FIG. 3 is a block diagram showing a configuration of a table data search method in an embodiment of the invention of claim 3;

FIG. 4 is a diagram showing an example in which table data is extracted from a database search result and recorded in table means.

FIG. 5 is a block diagram showing an embodiment including a table means having a title column set in the configuration of the table data search method in an embodiment of the invention of claim 3;

FIG. 6 is a diagram showing an example of implementation of a table in which column title columns are set.

FIG. 7 is a diagram showing an example of implementation of a table in which title columns of rows are set.

FIG. 8 is a diagram showing an example of implementation of a table in which title columns for both rows and columns are set.

FIG. 9 is a diagram showing an example of implementation of a table in which a title column in which item names form a hierarchy is set.

FIG. 10 is a diagram showing an implementation example of a generation rule for generating a description from an item name.

FIG. 11 is a diagram illustrating an example of implementation of the explanatory text of a table in which column title fields are set.

FIG. 12 is a diagram showing an embodiment of the explanatory text of a table in which both title columns of rows and columns are set.

FIG. 13 is a diagram showing an example of implementation of an explanatory note for performing a database search only on a specific area in a table.

FIG. 14 is a diagram showing an embodiment of an explanatory text of a table in which a title column in which item names form a hierarchy is set.

[Explanation of symbols]

 001 Table means 200 Natural language interface means 300 Database 101 Explanatory text input means 191 Search result extraction means 900 Positional information 901 Explanatory text 903 Item name 902, 907, 910 Search expression 911, 913, 915 Search result 912, 914, 916 Table data 003 Title column 102 Explanatory sentence generating means 905 Item information 906 Explanatory sentence generated from item name 103 Explanatory sentence recording means 104 Explanatory sentence synthesizing means 908 Recorded explanatory sentence 909 Synthesized explanatory sentence

[Procedure amendment]

[Submission date] June 14, 1994

[Procedure Amendment 1]

[Document name to be amended] Statement

[Correction target item name] Full text

[Correction method] Change

[Correction content]

[Document name] Statement

Title of table data retrieval device

[Claims]

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data input device used in a spreadsheet for tabulating data, and particularly for obtaining data to be input to the spreadsheet by searching a database existing independently of the spreadsheet. Table data retrieval device.

[0002]

2. Description of the Related Art Conventionally, for inputting data from a database into a spreadsheet, there has been an apparatus which directly reads a file storing data in the database and inputs it into a table for use. In order to directly read the file that stores the data, it is necessary to clarify the structure of the database and the format of the file that stores the beta.

Further, in order to search from a database input in a table, there has been an apparatus which inputs a search formula for a database and executes a database search. With this device as well, it is necessary to describe the search formula for database search after clarifying the structure of the database.

[0004]

In the conventional apparatus that directly reads the file of the database, it is difficult to selectively read the data because the format of the table at the time of reading is restricted by the storage format of the database. Later, the user had to format the table into the desired format.

Further, in the device for inputting the retrieval formula, it is necessary for the user to be able to describe the retrieval formula, and it is necessary to acquire the method of describing the database retrieval formula besides the method of using the spreadsheet.

Further, in any of the devices, there is a problem that the data cannot be used on the spreadsheet unless the details of the database are well known. Therefore, in order to use the data of the database on the tabulation spreadsheet. Has a problem that the user's burden is heavy.

An object of the present invention is to accept an input of a search sentence in a natural language from a spreadsheet, to execute a database search via a natural language interface for a database, in order to facilitate a database search from a spreadsheet. An object is to provide a tabular data retrieval device that allows the retrieval result to be input into a spreadsheet as tabular data.

[0008]

The gist of the means for solving the above problems is in the following three items.

[1] In a spreadsheet table data search device having a table means for recording data in a tabular format, an explanatory text input means for inputting an explanatory text relating to the data to be recorded in the table means in a natural language, Natural language interface means for converting the input explanatory text into a database search expression which is an instruction that the database can understand, and a search for performing a database search by the database search expression and transferring the result of the database search to the table means. A table data search device comprising: result extraction means.

[2] It is assumed that the table means has a title column for recording the item name corresponding to the data, the item name recorded in the title column is input, and when there are a plurality of input item names, these items are input. The table according to claim 1, further comprising: an explanatory sentence generating unit that generates an explanatory sentence regarding the data by combining the names, and the explanatory sentence generated by the explanatory sentence generating unit is input to the natural language interface unit. Data retrieval device.

[3] The explanation sentence recording means for recording the explanation sentence corresponding to the data recorded in the table means, and the explanation sentence recorded in the explanation sentence recording means are input, and the input explanation sentence is plural. And a description sentence synthesizing means for generating a synthesis explanation sentence in which these explanation sentences are combined, and the synthesis explanation sentence generated by the explanation sentence synthesizing means is input to the natural language interface means. The table data search device according to Item 1.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a table data search device of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a table data search method in an embodiment of the invention of claim 1.

The table data search device in this embodiment is
Display means 001 and natural language interface means 200
A database 300, an explanation input means 101,
The search result extracting means 191 is included.

A table data search device according to claim 1 will be described with reference to FIG. The explanatory text input means 101 is a natural language explanatory text 90 regarding the data to be recorded in the display means 001.
1 is received from the user and input to the natural language interface means 200. The explanatory note 901 may be a search command statement for searching the database 300 for data to be recorded in the table means 001.

The natural language interface means 200 can use, for example, the natural language interface method described in Japanese Patent Application No. 4-256660 filed by the applicant of the present invention, and inputs a description 901 in natural language such as Japanese, A search formula for searching the database 300 is generated.

For example, if the database 300 is a relational database and accepts SQL that is widely used as a search language for the relational database, the natural language interface means 200 is, for example, the IPSJ 43rd National Convention Lecture Collection 3- SQ configured to include the search expression generation method reported on pages 175 to 3-176.
The search expression 902 described in L is output.

The search result extracting means 191 is the database 3
If the search result 911 output from 00 includes information other than the data such as the number of searched data other than the one recorded in the table means 001, the search result 911 may be deleted or the search result 911 may be deleted. When the delimiter of the data 911 is different from the one recognized by the table means 001, the delimiter is converted and the search result 9 is obtained.
The table data 912 obtained by shaping 11 is recorded in the table means 001.

In FIG. 5, assuming that the database 300 contains information about company names and addresses,
An example in which a sentence “name and address of a company in M city” is assumed to be input as the explanatory sentence 901 relating to the table is shown. From the search result 911, the table data 912 obtained by converting a semicolon into a colon is input to the table means 001 except for the top line "There are 3 cases." When the table means 001 recognizes a colon as a delimiter, the table 001 is obtained.

FIG. 2 is a block diagram showing the arrangement of a table data search device according to the second embodiment of the invention. The table data search method of claim 2 will be described with reference to FIG.

In this embodiment, the table data search device has a table means 001 having a title column 003, an explanation sentence generating means 102, and a natural language interface means 200.
And a database 300 and a search result extracting means 191.

A title column 003 is set in the display means 001. The title column 003 is set to only columns as in the embodiment of FIG. 6, only rows as in the example shown in FIG. 7, or both rows and columns as in the example shown in FIG. It may be set. Further, the title columns of rows and columns may include item names 903 hierarchically.

FIG. 9 shows an embodiment in which the column item names 903 form two layers.

The explanation sentence generation means 102 uses the title column 003.
The item information 905 recorded in is input, and a description 906 for searching the database 300 for data corresponding to the item information 905 is output. The item information 905 includes an item name 903 and position information 900.

From the example of the title column shown in FIG.
For example, the sentence shown in FIG. 11 is generated. In the embodiment of FIG. 8, one of the item names 903 in the row title column and one of the item names 903 in the column title column is used as the search target, and the other is used as the search condition to generate the sentence shown in FIG. 12, for example. Also,
For example, when it is desired to perform a database search only on a specific area 801 in the table, for example, a sentence shown in FIG. 13 is generated from the item names 903 “profit” and “second half” related to the area 801. When the item names 903 are hierarchical as in the example shown in FIG. 9, the sentence shown in FIG. 14, for example, is generated by the conjunction of the hierarchical item names 903. The sentence generation from the item name 903 can be generated by a generation rule as shown in FIG. 10, for example.

Description 906 generated from item name 903
Is a search expression 9 by the natural language interface means 200.
It is converted into 07 and input to the database 300. The search result 913 based on the search expression 907 is the search result extraction means 19
The data is shaped into the table data 914 by 1 and is recorded in the position corresponding to the item information 905 in the table means 001.

FIG. 3 is a block diagram showing the arrangement of a table data search device according to the third embodiment of the invention. A table data search device according to claim 3 will be described with reference to FIG.

The table data search device in this embodiment is configured by adding an explanation sentence recording means 103 and an explanation sentence synthesizing means 104 to the table data retrieval device which is the embodiment of the invention of claim 1.

The description storage means 103 has a display means 001.
Record one or more explanations corresponding to the data recorded in. The explanation sentence synthesizing unit 104 receives one or more explanation sentences 908 output from the explanation sentence recording unit 103, and outputs the explanation sentence 909 obtained by combining the explanation sentences 908. The explanatory sentence 909 which is the output of the explanatory sentence synthesizing unit 104 is converted into a search expression 910 by the natural language interface unit 200 and input to the database 300. The search result 915 according to the search formula 910 is shaped into the table data 916 by the search result extracting means 191 and recorded in the table means 001.

Further, as shown in FIG. 4, the display means 001 may include a title column, and the explanatory note may be recorded in the explanatory note recording means 103 in association with the item name 903.

[0030]

According to the tabular data retrieval device of the present invention, an input of a retrieval sentence in a natural language is accepted from a spreadsheet, a database retrieval is executed through a natural language interface for a database, and the retrieval result is treated as tabular data. It is possible to input data into a spreadsheet and search the database from the spreadsheet.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of a table data search device according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a configuration of a table data search device according to an embodiment of the invention of claim 2;

FIG. 3 is a block diagram showing a configuration of a table data search device in an embodiment of the invention of claim 3;

FIG. 4 is a diagram showing an example in which table data is extracted from a database search result and recorded in table means.

FIG. 5 is a block diagram showing an embodiment including a table means in which a title column is set in the structure of the table data search device in one embodiment of the invention of claim 3;

FIG. 6 is a diagram showing an example of implementation of a table in which column title columns are set.

FIG. 7 is a diagram showing an example of implementation of a table in which title columns of rows are set.

FIG. 8 is a diagram showing an example of implementation of a table in which title columns for both rows and columns are set.

FIG. 9 is a diagram showing an example of implementation of a table in which a title column in which item names form a hierarchy is set.

FIG. 10 is a diagram showing an implementation example of a generation rule for generating a description from an item name.

FIG. 11 is a diagram illustrating an example of implementation of the explanatory text of a table in which column title fields are set.

FIG. 12 is a diagram showing an embodiment of the explanatory text of a table in which both title columns of rows and columns are set.

FIG. 13 is a diagram showing an example of implementation of an explanatory note for performing a database search only on a specific area in a table.

FIG. 14 is a diagram showing an embodiment of an explanatory text of a table in which a title column in which item names form a hierarchy is set.

[Explanation of Codes] 001 Table Means 200 Natural Language Interface Means 300 Database 101 Explanatory Text Input Means 191 Search Result Extraction Means 900 Position Information 901 Explanatory Text 903 Item Names 902, 907, 910 Search Expressions 911, 913, 915 Search Results 912, 914, 916 Table data 003 Title column 102 Explanatory sentence generating means 905 Item information 906 Explanatory sentence generated from item name 103 Explanatory sentence recording means 104 Explanatory sentence synthesizing means 908 Recorded explanatory sentence 909 Synthesized explanatory sentence

Claims (3)

[Claims] 1. A method for retrieving spreadsheet table data, comprising a table means for recording data in a database in a tabular form, comprising an explanatory sentence input means for inputting an explanatory sentence relating to the data to be recorded in the table means in natural language. , A natural language interface means for converting the input explanatory text into a command understandable by the database to generate a database search expression, and performing a database search by the database search expression, and transferring the result of the database search to the table means. And a search result extracting means for performing the table data search method. 2. The title means for recording the item name corresponding to the data is provided in the table means, and the item name recorded in the title column is input, and when there are a plurality of input item names, these item names are input. 2. The table data according to claim 1, further comprising: an explanatory sentence generation unit that generates an explanatory sentence relating to data by combining the explanatory sentences, and the explanatory sentence generated by the explanatory sentence generation unit is input to the natural language interface unit. retrieval method. 3. An explanation sentence recording means for recording an explanation sentence corresponding to data recorded in a table means, and an explanation sentence recorded in the explanation sentence recording means as an input, and a plurality of input explanation sentences. And an explanatory sentence synthesizing unit for generating a synthetic explanatory sentence in which these explanatory sentences are combined, and the synthetic explanatory sentence generated by the explanatory sentence synthesizing unit is input to the natural language interface unit. The table data retrieval method described in 1.